INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE THICK ASCENDING LIMB (TAL) of Henle's loop of the mammalian nephron reabsorbs ~30% of the NaCl filtered by the glomerulus and plays a crucial role in urinary concentrating ability. Basically, Na⁺ and Cl ions enter the cell via a Na⁺-K⁺-2Cl cotransporter at the apical membrane and, on the basolateral side, Cl ions leave the cell through a Cl conductance and a K-Cl cotransporter (Ref. 41; see also Fig. 9). In addition, a basolateral K⁺ conductance has been detected in the hamster medullary TAL (MTAL) (54), the rabbit cortical TAL (CTAL) (3), and the Amphiuma early distal tubule (13), the diluting segment in amphibians, and plays a key role in NaCl reabsorption by the TAL. This conductance would maintain basolateral membrane potential (V_Bl) difference and thus the driving force for the basolateral diffusion of Cl. In addition, by participating in the basolateral exit of K⁺ ions and compensating for part of the influx of K⁺ ions associated with Na⁺/K⁺-ATPase activity (24), this conductance is essential for the generation of a chemical gradient of Na⁺ ions promoting apical Na⁺-K⁺-2Cl cotransport activity.

The properties of ion channels underlying the basolateral potassium conductance of renal epithelial cells have been mainly studied in the proximal and cortical collecting tubules of various species (see Ref. 45). In contrast, very little is known about the conductive properties of the basolateral K⁺ channels in the TAL. The only patch-clamp study on this subject was restricted to channel properties in the cell-attached configuration and identified a 35-pS, inwardly rectifying K⁺ channel in rabbit CTAL (22).

The present study was undertaken to identify the basolateral potassium channels of mouse CTAL tubules. We were particularly interested in characterizing the conductive and regulatory properties of the channel that would allow a fruitful comparison with potassium channels described in other nephron segments and with cloned potassium channels.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals. EDTA (disodium or dipotassium salt, when appropriate), ATP (disodium salt), and N,N'-bis[3-aminopropyl]-1,4-butanediamine tetrahydrochloride (spermine; SPM) were from Sigma-Aldrich Chimie (Saint Quentin Fallavier, France). EGTA was from Research Organics (Cleveland, OH) or from Sigma-Aldrich Chimie.

Tubule preparation. Male 15- to 20-g CD1 (Charles River France, Saint Aubin Lès Elbeuf, France) or ICR (Harlan France, Gannat, France) mice were killed by cervical dislocation. CTALs were microdissected as previously described (14). Briefly, the left kidney was first rinsed with L-15 Leibovitz medium (Eurobio, Les Ulis, France; GIBCO BRL, Life Technologies, Cergy Pontoise, France) containing 300 U/ml collagenase (300 U/ml CLS II; Worthington, Freehold, NJ). Small pieces of cortex were then incubated at 37°C for 45-75 min in this collagenase-containing medium, rinsed, and kept at 4°C until microdissection. No further enzymatic or mechanical treatment of tubules was necessary for successful seal formation.

Current recordings. Cell-attached and cell-excised, inside-out variants of the patch-clamp technique (16) were applied to the basolateral membrane of CTAL fragments. Single-channel currents were recorded with a patch-clamp amplifier (LM-EPC 7, List Electronic, Darmstadt, Germany; RK-400, Bio-Logic, Claix, France) and stored on digital audiotape (DTR-1205, Bio-Logic). The bath reference was 0.5 M KCl in a 4% agar bridge connected to an Ag-AgCl pellet. In the cell-attached configuration, the clamp potential (V_c = V_bath  V_pipette) is superimposed on the cell membrane potential (V_m). In excised inside-out membrane patches, V_c = V_m. All V_c values were corrected for liquid-junction potential, as calculated by a routine of AxoScope software (Axon Instruments, Foster City, CA). The accuracy of these calculations for our experimental setup was confirmed by direct measurements utilizing a procedure described previously (30). Cations flowing from the inner to the outer face of the membrane patch are positive and shown as upward deflexions in current tracings. All experiments were conducted at room temperature.

Data analysis. Signals were generally low-pass filtered at 500 Hz by an eight-pole Bessel filter (LPBF-48DG; NPI Electronic, Tamm, Germany) and digitized at 3 kHz with a Digidata 1200 analog-to-digital converter and AxoScope software (Axon Instruments). For analysis of conductance substates (see RESULTS), signals were filtered at 2 KHz and digitized at 10 KHz. In most cases, channel activity was quantified using software kindly provided by Prof. T. Van den Abbeele (Paris, France). The mean current (I) passing through N channels was estimated from current-amplitude histograms and used to calculate the normalized current (NP_o) according to the equation NP_o = I/i, where i is the unit current amplitude. In some cases (see RESULTS), NP_o was measured with AxoScope software on the basis of visual differentiation between the fast and noisier K⁺ channel openings and the slower kinetics of 9-pS Cl channels (14, 15).

Channel selectivity. P_K/P_Na, the permeabilities to K⁺ and Na⁺ ions, respectively, was determined from both cell-attached and inside-out patches. In cell-attached patches, V_m was set close to 0 mV by superfusing the tubules with a high-K solution. The shift in reversal potential (E_rev) was then estimated from i-V_c relationships obtained with various K⁺ and Na⁺ concentrations in the pipette while the sum of K⁺ and Na⁺ concentrations was kept constant. P_K/P_Na was calculated using the Goldman-Hodgkin-Katz voltage equation (17), taking the E_rev in the presence of 144.8 mM K⁺ in the pipette (E_rev[144.8K]_pip) as a reference

where F is the Faraday constant, R is the universal gas constant, and T is absolute temperature. The use of this equation is based on the assumptions that channel anion permeability is negligible (see RESULTS) and that intracellular Na⁺ and K⁺ concentrations between cells are constant. P_K/P_Na on inside-out membrane patches was determined from E_rev values obtained with opposite Na⁺ and K⁺ ion gradients across the membrane patch, using the equation
where o and i denote the outside and inside faces of the membrane patch, respectively.

Solutions. Tubules were bathed either in a high-Na solution containing (in mM) 140 NaCl, 4.8 KCl, 10 glucose, and 10 HEPES-NaOH, pH 7.4, or in a high-K solution containing 144.8 KCl, 10 glucose, and 10 HEPES-KOH, pH 7.4. Free Ca²⁺ and Mg²⁺ concentrations were adjusted as required (see below). Unless otherwise stated, pipettes were filled with the high-K solution. Lower K⁺/Na⁺ concentration ratios were obtained by mixing appropriate volumes of high-K and high-Na solutions, both the sum of K and Na pipette concentrations ([K⁺]_pip + [Na⁺]_pip) and the Cl pipette concentration ([Cl]_pip) being kept at 144.8 mM. Higher values of [K⁺]_pip were obtained by adding the appropriate amount of KCl to the high-K solution. No correction was made for changes in osmolarity. When necessary (see RESULTS), ion currents through 9-pS Cl channels were minimized by using solutions in which all but 4.8 mM Cl had been replaced by NO (14). The pipette solution contained 1.2 mM free Mg²⁺. The bath solution typically contained 1.2 mM free Mg²⁺ and 1 mM free Ca²⁺. Various concentrations of free Mg²⁺ under low free Ca²⁺ concentrations were obtained with either EDTA or EGTA. The proportions of Mg²⁺ and EDTA or EGTA were calculated using the absolute stability constants for Mg²⁺, Ca²⁺, and H⁺ and EDTA (5) or EGTA (4, 5, 29).

Data presentation and statistics. Results are given as means ± SE for n patches. P values <0.05 (paired or unpaired t-test, when appropriate; SigmaPlot or SigmaStat, SPSS, Erkrath, Germany) were taken to represent statistically significant differences. Nonlinear regression analyses were performed with Origin software (Microcal Software, Northampton, MA). All-points amplitude histograms were constructed by TAC event detection and analysis software (Bruxton, Seattle, WA) from current traces after baseline current subtraction.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

K⁺ channel activity was observed in 34% (52 of 153) of cell-attached membrane patches of CTAL tubules bathed in the high-NaCl solution. With no applied potential, the number of active K⁺ channels per patch averaged 5.8 ± 0.5 (n = 52), with a mean NP_o of 1.9 ± 0.4 (range 0.02-10.49; n = 40) (Fig. 1A). We found no correlation between the tip resistance of KCl-filled pipettes (4-13 M) and the number of active channels (r = 0.017).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   K+ channel activity (NPo) in cell-attached patches. A: no. of patches displaying the indicated NPo at a clamp potential (Vc) = 4 mV. Tubules were bathed in high-Na solution, with the high-K solution in the pipette. B: typical recordings of K+ channel activity in the conditions given in A. Recordings were from the same patch at the Vc values given to the right of each trace. C, current level corresponding to closure of all K+ channels. Note the expanded scale for current amplitude obtained at +104 mV (*).

General properties of cell-attached patches. Figure 1B shows recordings of channel activity in a cell-attached patch from a tubule bathed in the high-NaCl solution, the pipette being filled with the high-KCl solution. The corresponding i-V_c relationship is shown in Fig. 2A. Under these conditions, E_K across the membrane patch should be close to 0 mV, and E_rev provides an estimate of V_m. E_rev was 75 ± 1 mV (n = 22), in good agreement with the V_m values of 70/80 mV reported for isolated perfused rabbit CTAL (12) and mouse MTAL (42) tubules. The i-V_c relationship for inward currents was nearly linear, single-channel G_in averaging 43.5 ± 0.6 pS (n = 22). Measurements of outward currents beyond E_rev were hampered by the openings of previously described small-conductance Cl channels (14) (not shown), but the current amplitude, measured on three occasions at 114 and 124 mV, was lower than expected for an ohmic K⁺ conductance (see Fig. 2A). The use of low-Cl solutions (see MATERIALS AND METHODS) facilitated measurements of K⁺ currents at more depolarizing voltages (not shown) and yielded an outward chord conductance [G_(chord)out], as measured between E_rev and 144 mV, of 11 ± 0.6 pS (n = 3), whereas G_in was 44 ± 2 pS (n = 9), indicating inward rectification.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   K+ channel properties in cell-attached patches. A: unit current amplitude (i)-Vc relationships in conditions of Fig. 1 () or in tubules bathed in a high-K solution (). Each point is the mean of 3-22 measurements, except at Vc = 124 mV, where n = 2. SE are error bars when larger than symbols. Inset: voltage dependence of NPo in conditions given in A (). Each point is the mean of 5-14 measurements, and error bars are SE. Dotted line, nonlinear least squares fit of Boltzmann equation (see text) to the data given by maximum NPo (NPo max) = 3.75, voltage when NPo is NPo max/2 (V1/2) = 11.5 mV, and effective valence (z) = 0.7 (r = 0.846). B: relationship between channel conductance and K+ concentration in the pipette ([K+]pip). For each [K+]pip value, the conductance at Vc = 0 mV (G0 mV) was taken as the slope of the i-Vc relationship between 10 and 10 mV in tubules bathed in high-Na solution. Each point is the mean of at least 3 determinations, and SE values are shown as error bars when larger than symbols. Dotted line, nonlinear least squares fit to mean values given by maximal G0 mV [G0 mV(max)] = 58 pS, dissociation constant (Km) = 83 mM, and value of G0 mV when [K+]pip = 0 [G0 mV(0)] = 7 pS (see text). C: ionic selectivity. The shift in reversal potential (Erev) was determined as described in MATERIALS AND METHODS from i-Vc relationships established in the cell-attached configuration in the presence of the indicated [K+]pip value. Each point is the mean of at least 4 determinations, and SE values are shown as error bars when larger than symbols. Dotted line through data points, nonlinear least squares fit using the Goldman-Hodgkin-Katz voltage equation (see MATERIALS AND METHODS) and a K-to-Na permeability ratio (PK/PNa) of 22.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Conductance substates of cortical thick ascending limb of Henle's loop (CTAL) K+ channels. A: excerpts of single-channel recordings from the same cell-attached patch. Tubule was bathed in high-Na solution, and patch was clamped at 0 mV. C, S1, S2, S3, and O: closed level, 1st, 2nd, and 3rd subconductance levels, and level of complete opening, respectively. B: histogram of current amplitude of each conductance substate, given as the fraction of the fully open level. S1, S2, and S3 were significantly different from each other (1-way ANOVA).

Channel conductive properties in cell-excised patches. As for basolateral K⁺ channels from rabbit CTAL (22), the activity of basolateral K⁺ channels from mouse CTAL was very rapidly lost on patch excision into physiological saline. However, as with other K⁺ channels, the rundown of which can be partly antagonized by using a Mg²⁺-free bath (35), we found that CTAL K⁺ channel activity could be maintained for up to 20 min by carrying out patch excision in a Mg²⁺- and Ca²⁺-free solution (+5 mM of either Na₂-EDTA or K₂-EDTA) in ~88% of patches. Results presented below were obtained according to this procedure.

Mechanisms underlying K+ channel inward rectification. Under symmetrical, high-K conditions, the i-V_c relationships in the absence of Mg²⁺ (+5 mM K₂-EDTA) (Fig. 4B) were linear. The mean values from four experiments were as follows: G_in and G_out were 50.5 ± 2 and 47.9 ± 2 pS, respectively (P = 0.41, paired t-test). The G_in/G_out ratio was 1.06 ± 0.06 (n = 4). We then investigated whether inward rectification in cell-attached patches resulted from a voltage-dependent block of outward currents by intracellular Mg²⁺ ions, as reported in other inwardly rectifying K⁺ channels (32, 37). Figure 4A compares the current amplitudes of an inside-out membrane patch, with its cytosolic side exposed to either an Mg²⁺-free medium (no Mg²⁺ + 5 mM K₂-EDTA) or to a medium containing 1.3 mM Mg²⁺ (1.5 mM MgNO₃ and 2 mM EGTA), at V_c of +70 mV. Internal Mg²⁺ reduced the single-channel amplitude of outward currents (Fig. 4A) and had little influence on inward currents (traces not shown). The corresponding i-V_c relationships are shown in Fig. 4B. Addition of 1.3 mM cytosolic Mg²⁺ slightly but significantly (P < 0.01) reduced G_in (42 ± 1.7 pS, n = 4, vs. 50.5 ± 2 pS, n = 4), but its most striking effect was a major reduction in G_(chord)out (measured between E_rev and +70 mV), which fell to 12.2 ± 1.58 pS (n = 4). The resulting G_in/G_out ratio was 3.8 ± 0.2 (n = 4). The i-V_c relationship for outward currents obtained under these conditions was very similar to that obtained from cell-attached patches on KCl-depolarized cells (compare with Fig. 2), and indeed the G_in/G_out ratios were not statistically different (P = 0.29, unpaired t-test).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Effects of internal Mg2+ and Na+ on i-Vc relationships of basolateral CTAL K+ channel. A: current traces from an inside-out membrane patch bathed in symmetrical high-K+ and the cytosolic side of which was bathed by either a medium containing 0 Mg2+ (0 Mg2++5 mM K2-EDTA; top trace) or a medium containing 1.3 mM Mg2+ (1.5 mM Mg2++2 mM EGTA; bottom trace). Clamped potential was +70 mV. C, current level corresponding to closure of all K+ channels. Also shown are corresponding all-point amplitude histograms from traces after baseline current subtraction (bottom). Solid lines are fits with a sum of 5 (0 Mg2+) or 3 (1.3 Mg2+) Gaussian components by the maximum-likelihood method. Means of components corresponding to channel opening level were 3.3 ± 0.75, 6.5 ± 0.74, 9.5 ± 0.8, 12.6 ± 0.79, and 16.5 ± 0.45 pA for 0 Mg2+ and 0.9 ± 0.21 and 1.9 ± 0.25 pA for 1.3 Mg2+. B: averaged i-Vc relationships corresponding to the conditions given in A for 0 () or 1.3 mM Mg2+ () or in the presence of a medium containing 0 Mg2+ (+5 mM K2-EDTA) supplemented with 20 mM Na+ on the cytosolic side of the patch (). Each point is the average of at least 3 determinations, and SE values are shown as error bars when larger than symbols. Solid line and , best fit to respective data, as given in RESULTS.

Intracellular SPM induces a voltage-dependent block. SPM is a naturally occurring polyamine that carries four positive charges at physiological pH and acts as a gating molecule, inducing a concentration- and voltage-dependent block of several inwardly rectifying K⁺ channels (6, 7, 26). We found that SPM rapidly and reversibly inhibited channel activity in cell-excised, inside-out membrane patches without affecting single-channel current amplitude (Fig. 5, insets). From Fig. 5, which compares the effects of 5 and 500 µM internal SPM on NP_o at both 60 and +60 mV, it is clear that the inhibitory effect of SPM was concentration and voltage dependent. This is quantified in Fig. 6A. At both membrane potentials, increasing SPM concentration reduced NP_o in a sigmoidal fashion. Fitting experimental data points with the Hill equation (see Fig. 6) revealed a channel sensitivity to SPM at positive potential (K_i = 0.2 µM) that was greater by two orders of magnitude than at negative potential (K_i = 26 µM). In contrast, Hill coefficients were similar and close to unity at both potentials (0.8 vs. 0.7, respectively). This voltage dependence was further investigated by observing the effects of 0.5 µM SPM on NP_o over an extended V_c range. The results of two separate experiments are illustrated in Fig. 6B. The relationship between NP_o and V_c was described by the Boltzmann distribution (Eq. 2). The best fit of data was given by NP_o max = 1.0, V_1/2 = 43 mV, and z = 1.25 (Fig. 6B).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Effects of intracellular spermine (SPM) on basolateral CTAL K+ channel activity. Traces show recording of K+ channel activity on the same cell-excised, inside-out membrane patch clamped at either 60 mV (top traces) or +60 mV (bottom traces). The patch was bathed in symmetrical high-K (144.8 mM, Cl- and Mg2+-free+5 mM Na2-EDTA). C, current level corresponding to closure of all K+ channels. Each trace was obtained in the absence (Control; left traces) or presence of 5 (middle traces) or 500 µM SPM (right traces) in the bath. Insets: a and b, excerpts at an expanded time scale (*) taken from the indicated sections of the traces.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Concentration and voltage dependence of the effect of internal SPM on basolateral CTAL K+ channel activity. A: dose-response relationships for the inhibitory effect of SPM on NPo at hyperpolarizing () and depolarizing () membrane potentials. The cytoplasmic side of cell-excised, inside-out membrane patches was exposed to the indicated SPM concentrations ([SPM]) at either Vc of 60 or +60 mV under the same conditions as in Fig. 5. Because of patch-to-patch variability, the effects of a given [SPM] on NPo were normalized to NPo for the same patch in the absence of SPM. Each value is the mean of at least 3 determinations, and SE are shown as error bars when larger than symbols. At both potentials, relationships between [SPM] and NPo were described by the equation NPo = 100/[1 + ([SPM]/Ki)nH], where Ki is the half-maximal inhibitory concentration, and nH is the Hill coefficient. The lines drawn are nonlinear, least squares fits to the respective data given by Ki = 26 ± 1 µM and nH = 0.8 ± 0.03 for hyperpolarizing voltages and Ki = 0.2 ± 0.06 µM and nH = 0.7 ± 0.1 for depolarizing voltages. B: voltage dependence of the effect of 0.5 µM SPM on NPo. Results obtained from 2 separate patches are shown ( and ). Within each experiment, NPo at a given Vc value in the presence of SPM was normalized to that for the same patch in the absence of SPM (control). Dotted line, is a nonlinear least squares fit of the Boltzmann equation (see RESULTS) to the data given by NPo max = 1.0, Vo = 43 mV, and z = 1.25.

Modulation by internal pH. Varying bath pH rapidly reversibly altered the K⁺ channel activity in cell-excised, inside-out membrane patches (Fig. 7A). NP_o was altered by intracellular pH (pH_i) in a sigmoidal fashion (Fig. 7B), and fitting data points using a modified Hill equation (Fig. 7) yield an apparent pK of 7.6. A Hill coefficient of 1.8 suggested that several protons were acting cooperatively in channel regulation by pH_i. The unit current amplitude through the CTAL K⁺ channel was not altered by pH_i. i-V_c relationships at pH 6.8 (n = 3) and 7.4 (n = 5) did not show any pH-dependent change in G_in (48 ± 4 vs. 45 ± 1 pS, respectively; P = 0.32) or G_out (36 ± 0.7 vs. 37 ± 3 pS, respectively; P = 0.83).

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Effects of variations in intracellular pH (pHi) on basolateral CTAL K+-channel activity. A: single-channel recordings from a cell-excised, inside-out membrane patch at the bath pH (i.e., pHi) values indicated on the right of each trace. Patch was bathed in symmetrical high-K (Cl- and Mg2+-free+5 mM Na2-EDTA) and clamped at Vc = 50 mV. C-, current corresponding to closure of all K+ channels. B: dose-response relationship of the effect of pHi on K+ channel activity under the same conditions as in A. Within each experiment, NPo at a given pHi value was normalized to NPo at pH = 7.4. Each point is the mean of at least 4 determinations, and SE values are shown as error bars when larger than symbols. The relationship between pHi and NPo is described by the equation NPo = NPo (max)/ [1 + ([H+]/K)nH], where NPo(max) is the maximal NPo, [H+] is the bath proton concentration, pK = logK, and nH is the Hill coefficient. The line drawn is a nonlinear least squares fit to the data given by NPo(max) = 2.93 ± 0.09, pK = 7.6 ± 0.1, and nH = 1.8 ± 0.1.

View larger version (27K): [in this window] [in a new window]   Fig. 8.   Effects of intracellular acidification on K+ channel activity in situ. The trace is a continuous recording of K+ channel activity of a cell-attached membrane patch with the high-Na solution in the bath. Pipette was filled with the high-K solution, and Vc was set at 0 mV. At the period indicated by the horizontal bar, 5 mM NH4Cl was added to the superfusate. C, current corresponding to closure of all K+ channels.

Internal ATP does not inhibit the basolateral CTAL K+ channel. ATP has no inhibitory effect on K⁺ channel activity (data not shown). Adding 1 mM internal ATP (Na salt) without Mg²⁺ for at least 1 min had no influence on channel activity of inside-out patches at V_c = 50 mV (P = 0.3, paired t-test, n = 3). No change in channel activity was seen when 5 mM internal ATP was added to the bath (n = 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Inward rectification properties. There has been very little investigation of rectification of K⁺ channels in native renal tissue (45). An intermediate rectification coefficient (G_in/G_out) of 4 has been explicitly determined for the basolateral 25-pS K⁺ channel in the Ambystoma proximal tubule (33). A G_in/G_out ~ 4-6 can also be extrapolated from published data on proximal tubule basolateral K⁺ channels (1, 20, 39). The i-V_c relationship of the basolateral CTAL K⁺ channel also displays a G_in/G_out of ~4.5 in cell-attached patches, indicative of an intermediate-type inward rectification. This differentiates the channel from strong inward rectifiers such as Kir 2.3 (CCD-IRK3) (48) and from the weak rectifier of the apical membrane of the cortical collecting duct (CCD) (47) and Kir 1.1 (2, 18) for which a G_in/G_out of 2-2.5 can be calculated. It is worth noting that the rectification would be hardly apparent in the presence of a physiological K⁺ concentration: G_in would be then comparable to G_out (~10 pS).

Channel regulation. The CTAL K⁺ channel is also sensitive to pH_i. A decrease in channel activity at acid pH is observed for many renal K⁺ channels (45), including the K⁺-secreting channel (and ROMK) in the apical membrane of the CCD and the proximal convoluted tubule basolateral channel (34, 35, 47). The pK of ROMK channels was 7.0-7.2, and the channel activity peaks at pH >7.4, as a consequence of a Hill coefficient of 3. A quite different pattern was observed for the CTAL channel, with a Hill coefficient of ~2 and a pK of ~7.6. Therefore, in situ channel activity would be in the lower part of the curve, so that even slight intracellular acidification would suffice to produce a significant reduction in NP_o. Our experiments indeed showed that acidifying the intracellular compartment by 0.3 pH unit was sufficient to inhibit channel activity in cell-attached patches. In view of the possible role played by this channel in NaCl reabsorption by the TAL (see below), its pH dependence is of physiological interest. Indeed, in vitro acute acidosis inhibits NaCl transport by the TAL (50). Thus, in addition to the basolateral pH-sensitive Cl channels (15), the 45-pS K⁺ channel we describe here appears as another basolateral target of acidosis in the inhibition of NaCl reabsorption by the TAL.

Comparison with other K+ channels in renal basolateral membranes. The only previous patch-clamp study in the basolateral membrane of CTAL reported the presence of a 35-pS K⁺ channel in the rabbit (22), showing inward rectification (G_out = 7 pS). As in the mouse, activity increases with depolarization in cell-attached patches. Thus the channels in rabbit and mouse CTALs seem to be similar, but some conductive and regulatory properties were not investigated in rabbit CTAL, which limits further comparisons. Several channels described in basolateral membranes of other nephron segments share some, but not all, properties of the CTAL channel. In the basolateral membrane of rat CCD, an inwardly rectifying K⁺ channel has been reported, but it is highly active and voltage independent, and pH insensitive and has a slightly lower conductance (27 pS) (28, 46). A 45-pS, pH-sensitive K⁺ channel has also been described in cell-attached patches of rat CCD (46), but it is activated by hyperpolarization. Other inwardly rectifying K⁺ channels have been studied in the basolateral membrane of the proximal tubule. Although some of these channels may differ in terms of voltage dependence, they do share a number of properties: they are all inhibited by acid pH (1, 34), and their inward conductance is ~50 pS in mammalian physiological saline (21, 38, 39) and depends on the external K⁺ concentration, with a K_m of 65-77 mM (23, 33). However, on the other hand, regarding inward rectification, the CTAL channel has a higher affinity toward Mg²⁺ and, unlike the PCT channel (33), exhibits high sensitivity to SPM. Second, the CTAL channel is not inhibited by ATP.

K+ conductance in the CTAL basolateral membrane. Studies of microperfused isolated tubules have given rise to conflicting interpretations on the issue of basolateral K⁺ conductance in CTAL. Greger and Schlatter (11) initially found that the basolateral membrane of rabbit CTAL was not conductive to K⁺ and postulated a basolateral K⁺ exit through a K⁺-Cl cotransporter (see Fig. 9). However, subsequent studies reported a K⁺-conductive pathway in the basolateral membranes of hamster and rabbit TAL (3, 54) and those of the Amphiuma diluting segment (13). The previous patch-clamp study in rabbit CTAL (22) and the present work in the mouse clearly establish that the basolateral membrane of CTAL cells is endowed with K⁺ channels.

View larger version (23K): [in this window] [in a new window]   Fig. 9.   Basic model of NaCl reabsorption by the TAL epithelium. Dotted lines indicate passive ion movements down electrochemical gradients. Vm, membrane potential.